Proceedings of the 3rd International Conference on Bigdata Blockchain and Economy Management, ICBBEM 2024, March 29–31, 2024, Wuhan, China

Research Article

Forecasting Financial Distress of Listed Firms Based on Recurrent Attention Networks

Download15 downloads
  • @INPROCEEDINGS{10.4108/eai.29-3-2024.2347392,
        author={Ming  Jia},
        title={Forecasting Financial Distress of Listed Firms Based on Recurrent Attention Networks },
        proceedings={Proceedings of the 3rd International Conference on Bigdata Blockchain and Economy Management, ICBBEM 2024, March 29--31, 2024, Wuhan, China},
        publisher={EAI},
        proceedings_a={ICBBEM},
        year={2024},
        month={6},
        keywords={self-attention; financial distress prediction; bert; text processing},
        doi={10.4108/eai.29-3-2024.2347392}
    }
    
  • Ming Jia
    Year: 2024
    Forecasting Financial Distress of Listed Firms Based on Recurrent Attention Networks
    ICBBEM
    EAI
    DOI: 10.4108/eai.29-3-2024.2347392
Ming Jia1,*
  • 1: Beijing Technology and Business University
*Contact email: jiaming9875@163.com

Abstract

Financial distress not only poses a threat to the long-term survival of a company, but also may have a chain reaction on the whole economic system. In recent years, the use of textual information as a feature of financial distress prediction has become a new hotspot, and this study proposes a financial text processing model based on recurrent attention network (RAN). Through an empirical study of A-share listed companies from 2007 to 2019, it is found that the RAN model performs well in extracting information from annual reports and effectively improves the accuracy of financial distress prediction.